Image2Count/cellevaluation.py at main · CancerTargetLab/Image2Count · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
import argparse
import os

def parse_args():
    parser = argparse.ArgumentParser(description="Evaluate results")

    # General arguments
    parser.add_argument("--vis_label_data", type=str, default="OC1_all.csv",
                        help="Count data of Images, linked with Patient IDs")
    parser.add_argument("--figure_dir", type=str, default="figures/",
                        help="Path to save images to")

    # Preprocess arguments
    parser.add_argument("--merge", action="store_true", default=False,
                        help="Wether or not to merge predictions of models in embed_dir")
    parser.add_argument("--embed_to_h5ad", action="store_true", default=False,
                        help="Wether or not to create an sc.AnnData obj and save it to out for predictions")
    parser.add_argument("--performance_metrics", action="store_true", default=False,
                        help="Wether or not to calculate performance metrics of stored sc.AnnData")
    parser.add_argument("--h5ad_dir", type=str, default="out/",
                        help="Path to h5ad object")
    parser.add_argument("--vis_name_pattern", type=str, default="model_[0-9]_all.h5ad",
                        help="Pattern of .5ad objects in 'h5ad_dir' which should be evaluated")
    parser.add_argument("--num_subgraphs_per_graph", type=int, default=0,
                        help="Number of subgraphs per graph to create to evaluate niches")
    parser.add_argument("--num_hops_per_subgraph", type=int, default=[], nargs='*',
                        help="Number of hops per subgraphs to create to evaluate niches")
    parser.add_argument("--sum_by_graph", action="store_true", default=False,
                        help="Wether or not 'vis_label_data' is label data or cell measurement data: ROI or Image key")
    parser.add_argument("--do_clustering_metrics", action="store_true", default=False,
                        help="Wether or not to calculate metrics which need to cluster predicted expression")
    parser.add_argument("--do_pathway_metrics", action="store_true", default=False,
                        help="Wether or not to calculate agreement between enricht CollecTRI, PROGENy and msigdb between true and actual counts base don true counts clusterin")
    parser.add_argument("--do_performance_metrics", action="store_true", default=False,
                        help="Wether or not to save metrics per markers/cell/niche")


    # Visualize Expression
    parser.add_argument("--visualize_expression", action="store_true", default=False,
                        help="Wether or not to visualize predicted sc expression")
    parser.add_argument("--has_expr_data", action="store_true", default=False,
                        help="Wether or not true Single Cell expression data is in measurements.csv")
    parser.add_argument("--raw_subset_dir", type=str, default="TMA1_preprocessed",
                        help="How the subdir in raw/ and processed/ is called, used when --has_expr_data is given")
    parser.add_argument("--processed_subset_dir", type=str, default="TMA1_preprocessed",
                        help="Subset directory of processed/ and raw/ of data")
    parser.add_argument("--embed_dir", type=str, default="out/",
                        help="Path to predicted single cell data per Graph/Image, if --merge is called this should be the parent directory containing all to be merged prediction folders, resulting in os.path.join(args['embed_dir'], 'mean') containing merged predictions")
    parser.add_argument("--vis_select_cells", type=int, default=0,
                        help="Number of cells to perform dim reduction on. If 0, then all cells get reduced")
    parser.add_argument("--vis_name", type=str, default="_cells",
                        help="Name added to figures name, saves processed data as NAME.h5ad")   #alo for visualize image

    # Visualize Image
    parser.add_argument("--visualize_image", action="store_true", default=False,
                        help="Wether or not to Visualize an Image")
    parser.add_argument("--vis_name_og", type=str, default="",
                        help="Name of NAME.h5ad containing TRUE single-cell counts(used for validation data)")
    parser.add_argument("--vis_img_raw_subset_dir", type=str, default="TMA1_preprocessed",
                        help="Name of raw/ subsetdir which contains .tiff Images to visualize")
    parser.add_argument("--name_tiff", type=str, default="027-2B27.tiff",
                        help="Name of .tiff Image to visualize")
    parser.add_argument("--figure_img_dir", type=str, default="figures/",
                        help="Path to output figures to")
    parser.add_argument("--vis_protein", type=str, default="",
                        help="Proteins to visualize Expression over Image of, seperated by ,; . converts to space")
    parser.add_argument("--vis_img_xcoords", nargs='+', type=int, default=[0,0],
                        help="Image x coords, smaller first")
    parser.add_argument("--vis_img_ycoords", nargs='+', type=int, default=[0,0],
                        help="Image y coords, smaller first")
    parser.add_argument("--vis_all_channels", action="store_true", default=False,
                        help="Wether or not to visualize all Image channels on their own")
    parser.add_argument("--size", type=int, default=25,
                        help="Size of dots as cells to be plotted")

    # Visualize Model Run
    parser.add_argument("--visualize_model_run", action="store_true", default=False,
                        help="Wether or not to Visualize statistics of model run")
    parser.add_argument("--model_path", type=str, default="out/models/ROI.pt",
                        help="Path and name of model save")
    parser.add_argument("--output_model_name", type=str, default="ROI",
                        help="Name of model in figures")
    parser.add_argument("--figure_model_dir", type=str, default="figures/",
                        help="Path to output figures to")
    parser.add_argument("--is_cs", action="store_true", default=False,
                        help="Wether or not Cosine Similarity is used or Contrast Loss")

    return parser.parse_args()


def main(**args):
    if args['merge']:
            from src.utils.utils import merge
            merge(args['embed_dir'])
            args['embed_dir'] = os.path.join(args['embed_dir'], 'mean')
    if args['embed_to_h5ad']:
        from src.utils.create_h5ad import createH5AD
        createH5AD(path=args['h5ad_dir'],
                   processed_dir=args['processed_subset_dir'],
                   embed_dir=args['embed_dir'],
                   label_data=args['vis_label_data'],
                   name=args['vis_name'])
    if args['performance_metrics']:
        from src.explain.Metrics import Metrics
        Metrics(path=args['h5ad_dir'],
                target=args['vis_label_data'],
                name=args['vis_name_pattern'],
                num_subgraphs_per_graph=args['num_subgraphs_per_graph'],
                num_hops_per_subgraph=args['num_hops_per_subgraph'],
                out=args['figure_dir'],
                sum_by_graph=args['sum_by_graph'],
                do_clustering_metrics=args['do_clustering_metrics'],
                do_pathway_metrics=args['do_pathway_metrics'],
                do_performance_metrics=args['do_performance_metrics'])
    if args['visualize_expression']:
        from src.explain.VisualizeExpression import visualizeExpression
        visualizeExpression(processed_dir=args['processed_subset_dir'],
                            embed_dir=args['embed_dir'],
                            label_data=args['vis_label_data'],
                            figure_dir=args['figure_dir'],
                            name=args['vis_name'],
                            select_cells=args['vis_select_cells'])
        if args['has_expr_data']:
            from src.explain.correlation import correlation
            correlation(raw_subset_dir=args['raw_subset_dir'],
                        figure_dir=args['figure_dir'],
                        vis_name=args['vis_name'])
    if args['visualize_image']:
        from src.explain.VisualizeImage import visualizeImage
        visualizeImage(raw_subset_dir=args['vis_img_raw_subset_dir'],
                       name_tiff=args['name_tiff'],
                       figure_dir=args['figure_img_dir'],
                       vis_name=args['vis_name'],
                       args=args)
    if args['visualize_model_run']:
        from src.utils.per_epoch_metrics import epochMetrics
        epochMetrics(model_path=args['model_path'],
                     figure_dir=args['figure_model_dir'],
                     is_cs=args['is_cs'],
                     name=args['output_model_name'])


if __name__ == '__main__':
    args = vars(parse_args())
    main(**args)